DESCRIPTION: (Applicant's Abstract) Violence has come to be recognized as a major public health issue. Research suggests a significant relationship between violence and drug/alcohol use among both victims and perpetrators, with notable ethnic, age, social context and gender differences. There also are indications that violence, as a force in community disruption and distrust, contributes to AIDS risk among substance abusers, and conversely, that fear of AIDS and the structure of gender relations in AIDS risk behavior, may contribute to interpersonal violence. Given the limited understanding of how violence interacts with and enhances substance use and AIDS risk behavior, especially in an understudied group like Puerto Ricans in the US, we seek to implement a baseline study of the relationships among violence, substance abuse, and AIDS. The proposed study will examine these relationships at the individual, social network, and wider social context levels. To accomplish this goal, the sample for this study comprising 400 individuals, will be divided into two groups; index and network participants. Index Participants will be 100 active, out-of-treatment Puerto Rican drug users. Index participants will be either IDUS or non-injection cocaine users. All index participants will be at least 18 years of age. They will be recruited through street outreach in four contrastive target neighborhoods, approximately 25% per neighborhood. Network Participants will be 300 individuals identified by index participants and members of their personal drug use and/or sexual networks. Network-based sampling of this kind, an approach called the random walk sampling technique, was developed and utilized in AIDS research by Klovdahl (1977,1985). The notable feature of this approach is that the index participant nominates a list of members of his/her network from which a random sample is selected for recruitment. All 400 participants will be interviewed at 4-month intervals for a twenty five month period using a battery of instruments that gauge participation in and experience of violence, drug use patterns, and AIDS risk behavior, as well as self-esteem, traumatic stress, and network characteristics. This longitudinal design will enable us to track the dynamic relations among the key variables over time. Additionally, a subsample of 50 index participants will be visited for an eight-hour period by a project ethnographer once each fourth cycle to allow: 1) a check on self-report; 2) opportunity to identify underlying influential connections between specific domains of behavior or groups of social actors in the complex social milieu in which violence, drug use, and AIDS interact; 3) observational information on behavior that helps to evoke more candid discussion of socially disapproved behavior; 4) an arena for detailed sequential study of behavior kinetics; 5) a means of studying the cognitive structures and organizing systems used by social actors to produce and pattern cultural behavior; and 6) serendipitous discovery of unknown or little understood behaviors. This approach will allow out team to develop a much clearer understanding of the relationship of violence to substance abuse and AIDS risk among Puerto Ricans, information needed to develop effective, culturally targeted prevention methods for a high risk population.